When Switching Algorithms Helps: A Theoretical Study of Online Algorithm Selection

TL;DR

This paper provides the first theoretical example demonstrating that switching between algorithms can asymptotically outperform individual algorithms, with a specific strategy achieving faster optimization times on OneMax.

Contribution

It introduces a theoretical framework showing how switching between two specific algorithms can lead to asymptotic speedups in optimization.

Findings

Switching algorithms can asymptotically outperform single algorithms.

A specific switching strategy achieves expected time $O(n ext{log} ext{log} n)$ on OneMax.

The analysis combines fixed-start and fixed-target perspectives.

Abstract

Online algorithm selection (OAS) aims to adapt the optimization process to changes in the fitness landscape and is expected to outperform any single algorithm from a given portfolio. Although this expectation is supported by numerous empirical studies, there are currently no theoretical results proving that OAS can yield asymptotic speedups (apart from some artificial examples for hyper-heuristics). Moreover, theory-based guidelines for when and how to switch between algorithms are largely missing. In this paper, we present the first theoretical example in which switching between two algorithms -- the $(1+\lambda)$ EA and the $(1+(\lambda,\lambda))$ GA -- solves the OneMax problem asymptotically faster than either algorithm used in isolation. We show that an appropriate choice of population sizes for the two algorithms allows the optimum to be reached in $O(n\log\log n)$ expected time, faster than the $\Theta(n\sqrt{\frac{\log n \log\log\log n}{\log\log n}})$ runtime of the best of these two algorithms with optimally tuned parameters. We first establish this bound under an idealized switching rule that changes from the $(1+\lambda)$ to the $(1+(\lambda,\lambda))$ GA at the optimal time. We then propose a realistic switching strategy that achieves the same performance. Our analysis combines fixed-start and fixed-target perspectives, illustrating how different algorithms dominate at different stages of the optimization process. This approach offers a promising path toward a deeper theoretical understanding of OAS.